Каталог навыков агентов

Claude Code toolkit - agents, commands, skills, rules, and hooks for productive AI-assisted development

Complete collection of Claude Code configs for C++20 HPC development - agents, skills, hooks, commands, and rules

Build the evaluation harness that gates every fine-tuning run — golden sets, per-failure-mode graders, judge calibration...

Build and run deterministic evaluation suites for agent workflows (single-turn or agentic). Use when you need reproducib...

Evaluate code quality, skill effectiveness, and implementation correctness against defined criteria with scoring and imp...

Score LLM and agent outputs using LLM-as-judge techniques — direct scoring against rubrics or pairwise comparison betwee...

Write LLM-as-judge evaluator prompts for AI products. Use this skill whenever you need to evaluate semantic properties o...

Use this after a task fails or produces wrong output. Documents the failure in growth-os/eval/corrections.md so future a...

Configure and run the isolated eval loop pattern — generate, evaluate, refine until pass threshold met

Improves prompt, skill, or instruction text in scored rounds against locked eval cases, with a frozen grader hash, repli...

Holdout-safe Outcomes grading: project the locked eval substrate, then ingest one verdict.

Orchestrator for the eval-output skill suite — evaluate LLM and agent outputs for quality, accuracy, helpfulness, and sa...

Guide for diagnosing and improving MSBuild project evaluation performance. USE FOR: builds slow before any compilation s...

Guide for diagnosing and improving MSBuild project evaluation performance. Only activate in MSBuild/.NET build context....

Design automated evaluation pipelines for LLM and agent systems — combining deterministic checks, statistical metrics, a...

Evaluate implementation plan before execution - validates architecture, coverage, dependencies, and best practices. TRIG...

Use when the user has run multiple evaluation skills and wants a comprehensive analysis — maturity assessment, cross-ski...

Generate evaluation report from session metrics, checkpoint evals, and pass@k data

Show a team skill's eval history — committed receipts per version and this machine's local runs — as a Markdown board fr...

Retroactively audit an executed AI phase's evaluation coverage — scores each eval dimension as COVERED/PARTIAL/MISSING a...

Write or update the eval rubric for an AI agent or AI feature: the pass/fail criteria that define good behavior, in four...

Design structured evaluation rubrics for scoring LLM and agent outputs — defining quality dimensions, scoring scales, ha...

Design a scoring rubric and LLM-as-judge prompt to evaluate the quality of an AI feature's output. Use when asked to cre...

Audit .claude/rules/ files for structural correctness, glob validity, and real-world usefulness. Resolves each paths: pa...

Run eval scenarios to benchmark Mycelium effectiveness. Execute tasks using reflexion loop, validate against success cri...

Audit all skills in the current project for frontmatter completeness, effort level appropriateness, allowed-tools scopin...

Eval and improve a skill against golden cases — run the target skill blind in a fresh, context-free subagent on each exa...

Batch evaluate multiple content pieces. Use when: scoring a content library, campaign assets, or deliverable set.

Develop and run agent behavior evaluations. Use this skill when asked to "write evals", "test agent behavior", "create e...

Rule matrix security fixture for eval_user used by security rule matrix integration tests; not intended for production d...

Designing and maintaining evaluations for LLM prompts, agents, and Agent Skills: success criteria, test-set composition,...

Guide for evaluation — inline evaluators, Inspect AI integration, and custom evaluators for measuring training progress....

Agent evaluation framework based on Anthropic's best practices. USE WHEN eval, evaluate, test agent, benchmark, verify b...

Use when evaluation results need triage and loop-closing — spec failures route to deterministic checks or context rules,...

Scaffold a first eval suite for an agent: mine real failures into cases, write behavioural checks over traces, and gener...

Use when draft eval criteria need refining, clustering, and acceptance into the published goldset with an isolated holdo...

Provides context about the Roo Code evals system structure in this monorepo. Use when tasks mention "evals", "evaluation...

Use when a published goldset is ready to become executable graders and configs — Python graders and metrics with auto-ru...

Use when standing up evals/{system}/ for the first time — scaffolds the EDD directory structure, picks PromptFoo or Deep...

Build and run evals for LLM and agent systems: golden datasets, LLM-as-a-judge with bias control, trajectory/step/outcom...

Use when you want guided bottom-up error analysis, structured failure taxonomy discovery, or comprehensive trace coding...

Use when a goldset with graders is ready to run — executes the evaluation pyramid and validates evaluator quality (SLA c...

Analyze a subject from five perspectives with assumption tracking and weighting. Use after research to examine all angle...

Structured UX evaluation that produces quantitative assessments, identifies specific issues, and routes to the right Int...

Evaluate an AI workflow against representative cases and its permitted actions. Use for model, retrieval or agent evalua...

Evaluate technologies and competitive developments against Genesis architecture

Evaluate AGV and AMR applicability from flow, payloads, routes, systems, traffic, safety, uptime, and capital intensity.

Evaluate AI security guardrails, moderation layers, prompt-injection detectors, policy engines, classifiers, and rule-ba...

Evaluate AS/RS applicability from SKU profile, cube, throughput, building constraints, storage height, automation, and c...

This SOP evaluates a completed brazil-bench attempt against the spec.md requirements, capturing metrics for comparison a...

Evaluates an ML GPU cluster for a cloud trial or acceptance test: environment dump, isolated newest PyTorch, matmul FLOP...

Evaluate conveyor applicability from throughput, product profile, dimensions, layout, labor, controls, safety, and capit...

Inspect software projects and evaluate application, database, and infrastructure deployment architectures across static...

Evaluate a generated diagram against a human reference using PaperBanana's VLM-as-Judge scoring.

計算資產價格相對長期指數成長趨勢線的偏離度，衡量當前是否處於歷史極端區間，並可選擇性地進行宏觀因子分析以判斷行情體質。

Critically assess external feedback (code reviews, AI reviewers, PR comments) and decide which suggestions to apply usin...

Evaluate and scope a free calculator, checker, generator, template, or interactive utility as a growth product. Use when...

Use when the user asks "should I bother applying", "is this worth it" or "should I still apply" about a saved job, or ne...

MMI評価エージェント - Modularity Maturity Indexによるモジュール成熟度評価（Cohesion/Coupling/Independence/Reusability）。/evaluate-mmi [対象パス] で呼...

Measure model performance on test datasets. Use when assessing accuracy, precision, recall, and other metrics.

Load the latest model checkpoint, run evaluation on the test set, and generate a metrics report with confusion matrix. U...

对物理题进行定量质量评估，判断题目是否适合用于学习、训练、讲义或命题。输出 100 分制评分、分项扣分理由、总体判断和修改建议。

How to verify the Cognitive OS is functioning correctly.

Load this skill when checking whether a plugin's skills, hooks, agents, or MCP servers work: write claude plugin eval ca...

Evaluate racking strategy options from load profile, selectivity, building constraints, MHE, throughput, and safety boun...

Design or audit an evaluation for retrieval-augmented generation, semantic search, or answer retrieval using a versioned...

Evaluate a single retort experiment run. Score the generated code against the task's TASK.md requirements, run its build...

Measure framework overhead by decomposing scenarios into clean-context forks and reporting per-fork load plus aggregate...

Validate and iterate on the SDLC Layer Separation Architecture implementation across the check categories defined in its...

Use when evaluating, filtering, scoring, validating, or pressure-testing a side project, startup idea, app idea, AI prod...

Evaluate a Claude Code skill or plugin you just wrote or changed, to find out whether it is any good before you ship it....

Use only when the user explicitly invokes evaluate-skill to compare a skill against a baseline or previous version using...

评价 Skill 执行结果的 Skill。当需要评价产出质量、判断是否需要迭代时触发。触发词：评价、evaluate、打分、怎么样、效果如何。

A fast, cheap sanity check on a Claude Code skill or plugin while you are still writing or iterating on it — not the ful...

Structured evaluation before adopting a library, tool, or hosted service — enumerate candidates wide, then score every o...

Evaluate one or more traces against an existing Truesight live evaluation. Use when a deployed live evaluation already e...

Help users make better hiring decisions. Use when someone is evaluating job candidates, making hiring decisions, conduct...

Make an evidence-based hiring decision and produce a Candidate Evaluation Decision Pack (criteria + scorecard, signal lo...

Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when ben...

Evaluates Node.js packages before installation by checking bundle size impact, comparing alternatives, verifying latest...

Scores whether a transformed image preserves the subject's identity (face / pose / composition stability) compared with...

Scores whether a produced image satisfies the original natural-language instruction. Use after any capability Skill prod...

Evaluate LLM systems using automated metrics, LLM-as-judge, and benchmarks. Use when testing prompt quality, validating...

Evaluate trained machine learning models with the right metrics and comparison logic. Use for benchmark review, threshol...

This skill allows AI assistant to evaluate machine learning models using a comprehensive suite of metrics. it should be...

Design an evaluation that proves an MCP agent's or server's quality with measurement rather than

Create a Technology Evaluation Pack (problem framing, options matrix, build vs buy, pilot plan, risk review, decision me...

Help users evaluate emerging technologies. Use when someone is assessing new tools, making build vs buy decisions, evalu...

Two-stage paper screening - abstract scoring then deep dive for specific data extraction

Evaluate skills by executing them across sonnet, opus, and haiku models using sub-agents. Use when testing if a skill wo...

Evaluates and selects Threat Intelligence Platform (TIP) products based on organizational requirements including feed in...

開発ツール、フレームワーク、ライブラリの評価と比較を支援します。PoC計画、複数候補の比較分析、意思決定フレームワークを提供します。技術選定、ツール導入の判断が必要な場合に使用してください。

Help users make better decisions between competing options. Use when someone is weighing pros and cons, comparing altern...

Evaluate trade-offs and produce a Trade-off Evaluation Pack (trade-off brief, options+criteria matrix, all-in cost/oppor...

Evaluate an OpenMed de-identification or clinical NER model against the leakage-first release gates G1a through G8, whic...